A scanner that measures what AI can read, and says so plainly
Search is becoming an answer rather than a list of links. ChatGPT, Claude, Gemini and Google's own AI answers now decide whether a business is seen, and that decision is made by machines reading pages, not by people browsing them. Lantad exists to measure one specific thing: whether those machines can actually read your site, and whether their answers name you.
What Lantad does
You paste a URL. Lantad fetches the page the way an AI crawler does, with no browser and no JavaScript, and compares that against what a real browser renders. It tests your robots.txt against each AI crawler token by name, checks whether your words survive the fetch, reads your structured data, and grades the result into an AI Visibility Score. The exact weights and the seven possible outcomes are published on the methodology page, and the behaviour of the crawler itself is documented on the crawler conduct page. The vocabulary it all rests on, from generative engine optimization to prose parity, is defined in the glossary.
The one rule that shapes everything
Lantad will not print a grade it could not earn. Every AI-visibility tool shows a number; when the measurement failed, that number is a guess wearing a lab coat. So Lantad's number refuses to appear: the report states what was measured, what was not, and why. That decision has its own write-up, why we will not grade a page we could not measure, and it is the clearest statement of how this tool differs from the category. A degraded scan that tells you the truth is worth more than a confident scan that is wrong.
It publishes the inconvenient findings
The evidence is public. Lantad runs the same scan across thousands of real sites and publishes what it finds, including findings that are awkward for the product. The crawlability study maps which platforms AI can and cannot read, the AI Visibility Index tracks the picture across the web, and the blog reports each measurement with the date it was taken and names what it did not measure. One of its own posts exists to say that llms.txt, a file the product ships a tool for, has no measured effect on the engines that matter. That willingness to publish against interest is the point, not an accident.
Who it is for
Lantad is built for the people who need to rank first in Google and be named by the AI chatbots at the same time: founders, in-house SEOs, and the agencies who answer to clients for both. It reads your pages as the crawlers those engines use, so you learn where you are invisible before a buyer does. If you want the fix as well as the finding, the done-for-you work is scoped from a scan rather than a sales call. You can see how to get cited across every answer engine, or just ask a named model what it already says about your brand.
Independent, and reachable
Lantad is independent and funded by the people who use it rather than by the companies it measures. It takes no money to move a grade, and the scoring weights are published so the grade can be checked against the code that produced it. If something here is wrong, or you want to talk about the work, email hi@lantad.co. The fastest way to understand what Lantad does is still to scan a page of your own and read the report.